The Agent Skills Directory

[SAFE]: Detailed analysis of all 10 threat categories confirms that the skill contains no malicious patterns, obfuscated code, or unauthorized data access attempts. The content is purely instructional and domain-specific.
[NO_CODE]: The skill consists solely of Markdown documentation and text-based communication templates. It does not include any scripts (Python, JavaScript, shell), executables, or configurations that would lead to dynamic code execution.
[PROMPT_INJECTION]: While the skill is designed to assist the agent in processing external data such as utility bills and interval data, the documentation itself does not contain any instructions to override system safety guidelines or bypass ethical protocols. The theoretical risk of indirect prompt injection from processed data is categorized as a low surface risk inherent to the domain.

energy-procurement